Differential Display RT-PCR Reveals Genes Associated with Lithium-Induced Neuritogenesis in SK-N-MC Cells

Abstract

Lithium is shown to be neurotrophic and protective against variety of environmental stresses both in vitro as well as in vivo. In view of the wider clinical applications, it is necessary to examine alterations in levels of expression of genes affected by lithium. Lithium induces neuritogenesis in human neuroblastoma cell line SK-N-MC. Our aim was to elucidate genes involved in lithium-induced neuritogenesis using SK-N-MC cells. The differential display reverse transcriptase polymerase chain reaction (DD-RT-PCR) technique was used to study gene expression profiles in SK-N-MC cells undergoing lithium-induced neuritogenesis. Differential expression of genes in control and lithium (2.5 mM, 24 h)-treated cells was compared by display of cDNAs generated by reverse transcription of mRNA followed by PCR using arbitrary primers. Expression of four genes was altered in lithium-treated cells. Real-time PCR was done to confirm the levels of expression of each of these genes using specific primers. Lithium significantly up-regulated NCAM, a molecule known to stimulate neuritogenesis, occludin, a molecule participating in tight junctions and PKD2, a molecule known to modulate calcium transport. ANP 32c, a gene whose function is not fully known yet, was found to be down-regulated by lithium. This is the first report demonstrating altered levels of expression of these genes in lithium-induced neuritogenesis and contributes four hitherto unreported candidates possibly involved in the process.

Introduction

Lithium is used as a drug for bipolar disorder and is likely to find wider applications in treatment for neurodegenerative diseases (Chen et al. 2000; Gould and Manji 2002; Nowicki et al. 2008) as it is shown to be neuroprotective (Nowicki et al. 2008; Zhang et al. 2005). It protects neurons against variety of environmental stresses both in vitro as well as in vivo (Quiroz et al. 2004; Yazlovitskaya et al. 2006; Yuan et al. 1998). Lithium competes with magnesium (Quiroz et al. 2004; Ryes and Harwood 2001) and at therapeutic doses inhibits enzymes of second messenger systems (Chen et al. 1999, 2000; Manji and Chen 2002; Yuan et al. 1998) or those key to regulating cell death or survival (Quiroz et al. 2004; Zhang et al. 2005). Lithium is also known to be neurotrophic and is reported to enhance neuritie extensions and synaptogenesis (Quiroz et al. 2010). In view of the potential applications, it is necessary not only to elucidate its exact mechanism of action but also to examine alterations in levels of expression of genes affected by lithium in other pathways. Short- or long-term exposure of lithium and its influence on gene expression has been investigated in vivo (Bosetti et al. 2002; McQuillan et al. 2007) and in vitro (Brandish et al. 2005; Seelan et al. 2008; Zhang et al. 2005). Study of gene expression profiles by cDNA microarrays of rat brain after 7 (sub-acute) and 42 days (chronic) of oral lithium administration demonstrated that participating genes under these two cellular conditions were different (Bosetti et al. 2002). Another study using same technique on brain cells of mice fed with lithium for 2 weeks identified 121 genes that were significantly regulated under these conditions (McQuillan et al. 2007). A report on early gene expression profiles of lithium-induced apoptosis in human juvenile costal chondrocytes derived T/C28a cells has listed 50 differentially expressed genes by cDNA micro array (Zhang et al. 2005). A major transcriptome analysis of lithium-treated human neuroblastoma cell line SK-N-AS reported differential expression of several unidentified genes in addition to known genes involved in signaling pathways (Seelan et al. 2008). Unlike these reports, we have used DD-RT-PCR technology and identified four more genes not reported in any of the previous studies on lithium action.

We used human neuroblastoma cell line SK-N-MC as model system to identify genes that showed variation in levels of expression after 24 h of treatment with lithium. Lithium induces neuritogenesis in human neuroblastoma cell line SK-N-MC. Differential expression of genes was elucidated by comparing differentially displayed cDNAs generated by reverse transcription of mRNA from lithium-treated SK-N-MC cells followed by PCR with arbitrary primers and re-confirmation by real-time (RT) PCR using gene specific primers. Here, we report, altered expression of hitherto unreported genes associated with lithium-induced neuritogenesis for the first time, thereby suggesting possible new mechanisms of its action.

Methods

Cell Culture and Treatment

Human neuroblastoma cell line SK-N-MC (no. HTB-10) was obtained from American Type Culture Collection, USA, and maintained in DMEM (GIBCO) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (0.75 μg/ml). Cells were treated with 2.5 mM lithium chloride prepared in FBS supplemented DMEM for 24 h. Chemicals used were from Sigma, USA, if not otherwise stated.

Cell Viability and Morphology

Effect of lithium on cell viability was measured by MTT assay. For morphological observations, cells seeded (7,500 cells/100 μl) on cover slips, with or without lithium (2.5 mM) for 24 h were fixed with methanol:acetic acid (4:1), stained with Giemsa, and observed for number per cells bearing more than three neurites. Images of SK-N-MC (≥30) from randomly selected fields were obtained. Cytoplasmic extension of ≥0.1 mm from the soma was considered as a neurite. The length of the neurite was taken as the distance from the soma to the tip of the neurite (Kim 2005) and quantified using Image J software.

RT-PCR and Differential Display of cDNA

SK-N-MC cells were harvested with chilled Trizol (Ambion, USA) directly in the culture flasks and followed up for RNA extraction as per manufacturer’s instructions. Re-extraction with Trizol was done after treatment of samples with RNase-free DNase (Promega, USA). RNA (5 μg) from control- and drug-treated cells was reverse transcribed using 2 base anchored oligodT reverse primer and enzyme from Revert Aid First Strand cDNA synthesis kit (MBI Fermentas, USA).

Differential Display Reaction

Second-strand cDNA synthesis was initiated with 200 ng of cDNA using arbitrary primers BRNS 2 (5′-CTT GAT TGC C-3′)/BRNS 4 (5′-CGA AGC GAT C-3′) and anchor primer (5′-TGC CGA AGC TTT TTT TTT TTT GC-3′). All primers were generated using DNAMAN software. Cycling profile for amplification by polymerase chain reaction (PCR) was 94°C (2 min), low stringency annealing at 40°C (2.30 min), and extension at 72°C (2.30 min) for 3 cycles followed by 30 cycles of 94°C (30 s), 40°C (2 min), 72°C (1 min + 4 s/cycle), with final extension of 7 min at 72°C. RT-PCR products from control- and lithium-treated RNA samples were analyzed on consecutive lanes of sequencing sized 5% denaturing polyacrylamide Tris–borate–EDTA gels following electrophoresis at 1,800 V. Bromophenol blue marker dye front was allowed to run out of gels. Gels were stained with silver nitrate following fixation in methanol (2 × 20 min). Differentially expressed fragments of cDNAs were carefully excised using sterile scalpel and re-synthesized by PCR using corresponding arbitrary and anchor primers. Higher stringency of annealing temperature was used to ensure synthesis of homogenous amplicons (ones with maximum complementarity to primers). The cycling parameter was one cycle of 94°C (2 min), followed by 94°C (1 min), 50°C (1 min), 72°C (1 min) for 30 cycles and 12 min final extension at 72°C, to aid generation of template independent extensions of the amplicons for direct cloning.

Cloning and Identification of Genes

All re-amplified cDNA products were cloned in T vector (MBI Fermentas, USA) and maintained in E. coli DH5α cells. Positive clones were sequenced on a 377-18 ABI Prism automated DNA sequencer, using vector primers. DNA sequences were searched using BLAST in NCBI nucleotide database.

Real-Time PCR Using SYBR-Green Chemistry

Real-time quantitative PCR (20 μl) amplification reactions were carried out in an Eppendorf Master Cycler EP Realplex sequence detection system. The reaction mixture consisted of 9 μl 1× PCR Master mix containing SYBR-Green, 2 pmol of each gene specific primers, and cDNA template. To determine the unregulated endogenous reference genes in the samples, Ct-values of β-actin and 18S rRNA were tested simultaneously. In this case, Ct-value was the cycle number at which the fluorescence generated within a reaction crosses the threshold within the linear phase of the amplification profile. The Ct-value is an important quantitative parameter in real-time PCR analysis (Bernard and Wittwer 2002; Florl et al. 2005). All reactions were carried out in triplicates with appropriate control reactions. The PCR conditions were: 95°C for 2 min, followed by 40 cycles of 95°C for 30 s, and 60°C for 30 s and 72°C for 30 s. The fluorescent spectra were recorded during the elongation phase of each PCR cycle. The Ct-values were calculated with sequence-detection system (SDS) software V1.7 (Applied Biosystems) and an automatic setting of base line. The relative quantification was performed as outlined in the Relative Quantification of Gene Expression: (Applied Biosystems). A comparative Ct method was used (Gutala and Reddy 2004). Briefly, this comparative Ct method involved averaging triplicate samples taken as the Ct-values for gene of interest and β-actin. The ΔCt-value was obtained by subtracting the average β-actin Ct-value from the average Ct-value of gene of interest. The fold change was calculated according to the formula 2^−(ΔΔCt), where ΔΔCt was the difference between ΔCt of sample and the ΔCt calibrator.

Results

Transcriptome of any living cell is a dynamic entity and it is very important to maintain experimental integrity during analysis of the same using any of the existing low or high throughput techniques. Care was taken to maintain same culture conditions and treatments parameters for the SK-N-MC cells for each DD-RT-PCR experiment. No change in cell viability compared to control cells (no lithium) was seen upon treatment with lithium for 24 h (Table 1) suggesting that lithium was non-toxic at this dose. Although viability was unaffected, morphology of cells was altered due to lithium exposure. This was confirmed by the observation that percentage of cells bearing three or more neurite extensions was significantly increased in treated cells compared to that of controls (Table 1). Lithium in addition to being neuroprotective is known to promote new neurite sprouting and enhance neurite outgrowth in in vitro models (Chen and Manji 2006). In the present system, at an early time point, lithium seems to have induced neuritis while no alteration in neurite length was observed. Lithium is known to activate extracellular signal regulated pathways that mediates its neurotrophic action (Chen and Manji 2006). Molecular mechanism that triggers neurite outgrowth and those that co-ordinate changes in cytoskeleton that result in neurite extension are not clearly understood and could be different or temporarily separated. For example, NCAM, a molecule reported to be up-regulated by lithium in the present study, in other models is known to trigger neurite outgrowth (Kleene et al. 2010; Williams et al. 1992) but there are no reports explicitly implicating NCAM in neurite elongations. We have therefore used this model to elucidate gene involved essentially in neurite sprouting. Henceforth, cells treated with lithium (2.5 mM) for 24 h were used for all experiments to identify genes associated with lithium-induced neuritogenesis in SK-N-MC cells. Four reproducible cDNA fragments from consecutive DD-RT-PCR experiments were followed up for sequence identification wherein homology search in NCBI database showed 98% match to genes neural cell adhesion molecule (NCAM), polycystic kidney disease interactor 2 (PKD2), occludin and acidic nucleo phosphoprotein 32c (ANP32c), respectively. Figure 1 depicts real-time PCR profiles of these four genes that were differentially expressed under the present experimental condition. The amount of target cDNA in the control and treated samples was compared after PCR amplification by fluorescence monitoring of each cycle. Early appearance of crossing threshold (Ct) for occludin, PKD2, and NCAM cDNAs (Fig. 1a–c) in lithium-treated sample indicated their up-regulation against control. In contrast, for ANP32c (Fig. 1d), the Ct appears later than control suggesting down-regulation of this gene. Thus, the obtained Ct values confirmed differential expression of each cDNA in control- and lithium-treated cells. This reinforced our original observation of differentially silver stained cDNA fragments on the DD-RT-PCR gels (data not shown). Brief descriptions about each of these genes and their chromosomal locations have been provided in Table 2 for reference.

Table 1.

Viability and morphological parameters of SK-N-MC cells treated with lithium (2.5 mM, 24 h)

	Percent viability (MTT assay) Mean ± SD 5000 cells/well	Neurite length/Cell (mm) Mean ± SD	Percent cells bearing ≥3 neurites
Control	100	0.23 ± 0.09	14.9
Lithium	105.40 ± 1.72	0.21 ± 0.07	21.2*

* Significant at P > 0.05, (n = 30)

Fig. 1.

Real-time PCR profiles (in duplicates) of cDNA amplicons from control (ctrl), lithium (Li)-treated SK-N-MC cells, and housekeeping gene β-Actin. SYBR Green fluorescent intensities in arbitrary units (Y-axis) with an increasing number of thermal cycles (X-axis) was observed by real-time PCR cycles in each profiles. The profiles depicted are a occludin, b PKD2, c NCAM, and d ANP-32c

Table 2.

Differentially expressed genes in lithium (2.5 mM, 24 h) treated SK-N-MC cells

Gene name	Chromosome location	Fold Change	Accession no.	Function
ANP-32c	4	−2.67 ↓	NW_92217	Leucine-rich protein homologous to ANP-32a. It is the mutant of the same and is known to be tumorogenic (Florl et al. 2005; Muresan et al. 2000)
NCAM	21	19.55 ↑	NW_927384	Plays an important role in development, synaptic plasticity, and regeneration (Bernard and Wittwer 2002; Kleene et al. 2010)
Occludin	5	4.48 ↑	NW_922784	Important in formation of tight junction, BBB, and aqueous pores. Plays a role in diabetes, inflammation, and cancer (Feldman et al. 2005; Kular et al. 2010; Seelan et al. 2008)
PKD2	22	4.92 ↑	NW_927628	Involved in polycystin kidney disease. Known to modulate calcium influx (Kochevar et al. 2004; Newby et al. 2002)

Discussion

Lithium, at 24 h, down regulated mRNA expression of ANP-32c in SK-N-MC cells. ANP32c or ppr32-1 belongs to a family of highly conserved leucine-rich acidic nucleoproteins (LANP) family of genes (Kular et al. 2010). The other well-known member of this family of gene is ANP32b (pp32), known to be associated with inhibition of histone acetylation and is an inhibitor of protein phosphatase 2A (Munemasa et al. 2008). While there is a degree of protein homology between ANP32a and ANP32c, functionally they appear divergent as ANP32a is known to be tumor suppressor, whereas ANP32c is proto oncogene though little is known about its function (Kochevar et al. 2004). It has been reported that ANP32a is expressed predominantly in nucleus in undifferentiated neuronal cells, while it translocates to cytoplasm during neuritogenesis and interacts with microtubule associated protein 1B (MAP1B) (Muresan et al. 2000; Opal et al. 2003). We report that ANP32c unlike ANP32b is decreased during lithium-induced neuritogenesis. This suggests that ANP32c and ANP 32a may have divergent function in neuritogenesis as in tumorogenecity and needs to be investigated further.

Lithium at 24-h up regulated mRNA expression of three molecules in SK-N-MC cells, and interestingly, all of these are expressed on the cell surface. Lithium robustly up-regulated NCAM (≈20 times) neural adhesion molecule, known for neural cell differentiation and neuritogensis by interacting with other NCAM molecules on adjacent cells (Kolkova et al. 2000). Increase in expression of occludin during neuritogenesis is a novel finding as so far this protein was only known to be a component of tight junctions in epithelial cells (Feldman et al. 2005; Muresan et al. 2000). Only recently another group of tight junction proteins have been implicated in neurodegerative diseases (Romanitan et al. 2010) and our observation with occludin deserves attention for possible role of this molecule in similar novel functions. Up-regulation of the third mRNA in our study of the calcium channel gene, PKD-2, implicated in polycystic kidney disease (Ma et al. 2005; Newby et al. 2002) has similar possibilities as no functions have been reported in neuronal cells yet. None of these molecules have been implicated in lithium action in neuronal cells before. Interestingly though, all of them are directly or indirectly modulated by signaling pathways known to be altered by lithium action.

In conclusion, we report here a model system for neuritogenesis in SK-N-MC cells, to study lithium-induced alterations of levels of four genes. Expression of occludin, a molecule participating in tight junctions, PKD2 a molecule involved in calcium channel and NCAM, a molecule known to stimulate neuritogenesis, were significantly up regulated, while expression of ANP32c, (function unknown) was down regulated. None of these genes were reported to be modulated in studies with short-term or chronic treatment of lithium and could well be associated with lithium-induced neuritogenesis. At this point, it is difficult to comment whether this observation is universal or restricted to our model only (i.e. SK-N-MC cells) though it adds to the gene list under lithium effect. Further studies with knock down models of each of these genes in this cell line could be useful from therapeutic point of understanding neuritogenesis.